A global meta-analysis of microarray expression data to predict unknown gene functions and estimate the literature-data divide

Clover: An unbiased method for prioritizing differentially expressed genes using a data-driven approach

Identifying key genes from a list of differentially expressed genes (DEGs) is a critical step in transcriptome analysis. However, current methods, including Gene Ontology analysis and manual annotation, essentially rely on existing knowledge, which is highly biased depending on the extent of the literature. As a result, understudied genes, some of which may be associated with important molecular mechanisms, are often ignored or remain obscure. To address this problem, we propose Clover, a data-driven scoring method to specifically highlight understudied genes. Clover aims to prioritize genes associated with important molecular mechanisms by integrating three metrics: the likelihood of appearing in the DEG list, tissue specificity, and number of publications. We applied Clover to Alzheimer's disease data and confirmed that it successfully detected known associated genes. Moreover, Clover effectively prioritized understudied but potentially druggable genes. Overall, our method offers a novel approach to gene characterization and has the potential to expand our understanding of gene functions. Clover is an open-source software written in Python3 and available on GitHub at https://github.com/G708/Clover.

BioVDB: biological vector database for high-throughput gene expression meta-analysis

High-throughput sequencing has created an exponential increase in the amount of gene expression data, much of which is freely, publicly available in repositories such as NCBI's Gene Expression Omnibus (GEO). Querying this data for patterns such as similarity and distance, however, becomes increasingly challenging as the total amount of data increases. Furthermore, vectorization of the data is commonly required in Artificial Intelligence and Machine Learning (AI/ML) approaches. We present BioVDB, a vector database for storage and analysis of gene expression data, which enhances the potential for integrating biological studies with AI/ML tools. We used a previously developed approach called Automatic Label Extraction (ALE) to extract sample labels from metadata, including age, sex, and tissue/cell-line. BioVDB stores 438,562 samples from eight microarray GEO platforms. We show that it allows for efficient querying of data using similarity search, which can also be useful for identifying and inferring missing labels of samples, and for rapid similarity analysis.

The systematic comparison between Gaussian mirror and Model-X knockoff models

While the high-dimensional biological data have provided unprecedented data resources for the identification of biomarkers, consensus is still lacking on how to best analyze them. The recently developed Gaussian mirror (GM) and Model-X (MX) knockoff-based methods have much related model assumptions, which makes them appealing for the detection of new biomarkers. However, there are no guidelines for their practical use. In this research, we systematically compared the performance of MX-based and GM methods, where the impacts of the distribution of explanatory variables, their relatedness and the signal-to-noise ratio were evaluated. MX with knockoff generated using the second-order approximates (MX-SO) has the best performance as compared to other MX-based methods. MX-SO and GM have similar levels of power and computational speed under most of the simulations, but GM is more robust in the control of false discovery rate (FDR). In particular, MX-SO can only control the FDR well when there are weak correlations among explanatory variables and the sample size is at least moderate. On the contrary, GM can have the desired FDR as long as explanatory variables are not highly correlated. We further used GM and MX-based methods to detect biomarkers that are associated with the Alzheimer's disease-related PET-imaging trait and the Parkinson's disease-related T-tau of cerebrospinal fluid. We found that MX-based and GM methods are both powerful for the analysis of big biological data. Although genes selected from MX-based methods are more similar as compared to those from the GM method, both MX-based and GM methods can identify the well-known disease-associated genes for each disease. While MX-based methods can have a slightly higher power than that of the GM method, it is less robust, especially for data with small sample sizes, unknown distributions, and high correlations.

Identification of novel potential molecular targets associated with pediatric septic shock by integrated bioinformatics analysis and validation of in vitro septic shock model

Background/Aim: Sepsis is a major cause of morbidity, mortality, and healthcare utilization among children all over the world. Sepsis, characterized as life-threatening organ failure, results from a dysregulated host response to infection. When combined with critically low blood pressure, it causes septic shock, resulting in high mortality  rates. The aim of this study was to perform a bioinformatic analysis of gene expression profiles to predict  septic shock risk. Methods: Four datasets related to pediatric septic shock were retrieved from the Gene Expression Omnibus (GEO) database for a total of 240 patients and 83 controls. GEO2R tools based on R were used to find differentially expressed genes (DEGs). The Database for Annotation, Visualization and Integrated Discovery (DAVID) was used to examine the functional enrichment of DEGs. STRING was used to create a protein–protein interaction (PPI) network. After separately analyzing the four datasets, commonly affected genes were removed using the Venny program. Finally, human umbilical vein endothelial cells (HUVECs) were stimulated with supernatants of lipopolysaccharide (LPS)-stimulated RAW267.4 macrophage cells and expression of selected genes was confirmed by real-time reverse-transcriptase polymerase chain reaction (qRT-PCR) and used to construct an in vitro septic shock model. Results: Seven-hundred seventy-one common differentially expressed genes in the four  groups were found. Of these, 433 genes showed increased expression, while 338 had reduced expression. In the DAVID analysis results, DEGs up-regulated according to gene ontology results were enriched in the regulation of innate and adaptive immune responses, complement receptor-mediated signaling, and cytokine secretion processes. Down-regulated DEGs were  significantly enriched in the regulation of immune response, T-cell activation, antigen processing, and presentation and integral component of plasma membrane processes. According to The Search Tool for the Retrieval of Interacting Genes/Proteins (STRING), Cystoscape Molecular Complex Detection (MCODE), nine down-regulated genes in the center of the PPI network, ZAP70, ITK, LAT, PRKCQ, LCK, IL2RB, FYN, CD8A, CD247 and four up-regulated genes, MMP9, TIMP1, LCN2, HGF, were associated with septic shock. Expressions of FYN and MMP9 genes in the in vitro septic shock model were consistent with the bioinformatic results. Conclusion: Comparative bioinformatics analysis of data from four different septic shock studies was performed. As a result, molecular processes and important signal networks and 13 genes that we think will play a role in the development and risk prediction of septic shock are proposed. Methods: Four datasets related to Pediatric septic shock were retrieved from the Gene Expression Omnibus (GEO) database for a total of 240 patients and 83 controls. GEO2R tools based on R were used to find differentially expressed genes (DEGs). DAVID was used to examine the functional enrichment of DEGs. STRING was used to create a protein-protein interaction (PPI) network. After separately analyzing the four datasets, commonly affected genes were removed using the Venny program. Finally, HUVECs were stimulated with supernatants of LPS-stimulated RAW267.4 macrophage cells and expression of selected genes was confirmed by qRT-PCR, constructing an in vitro septic shock model. Results: There were 771 common differentially expressed genes in the 4 groups. Of these, 433 genes showed increased expression, while 338 had reducing expression. In the DAVID analysis results, DEGs upregulated by gene ontology were enriched in the regulation of innate and adaptive immune responses, complement receptor-mediated signaling, and cytokine secretion processes. Downregulated DEGs are significantly enriched in the regulation of immune response, T cell activation, antigen processing, and presentation and integral component of plasma membrane processes. According to STRING, cystoscape MCODE, and cytohubba analysis, 9 downregulated genes in the center of the PPI network, ZAP70, ITK, LAT, PRKCQ, LCK, IL2RB, FYN, CD8A, CD247, and 4 upregulated genes, MMP9, TIMP1, LCN2, HGF, were associated with septic shock. Expressions of FYN and MMP9 genes in the in vitro septic shock model were consistent with bioinformatic results. Conclusion: Important signaling networks and 13 genes potentially indicating molecular processes for the incidence, development, and risk prediction in septic shock were found using bioinformatic analysis of gene expression profiles.

Oklahoma Nathan Shock Aging Center - assessing the basic biology of aging from genetics to protein and function

The Oklahoma Shock Nathan Shock Center is designed to deliver unique, innovative services that are not currently available at most institutions. The focus of the Center is on geroscience and the development of careers of young investigators. Pilot grants are provided through the Research Development Core to junior investigators studying aging/geroscience throughout the USA. However, the services of our Center are available to the entire research community studying aging and geroscience. The Oklahoma Nathan Shock Center provides researchers with unique services through four research cores. The Multiplexing Protein Analysis Core uses the latest mass spectrometry technology to simultaneously measure the levels, synthesis, and turnover of hundreds of proteins associated with pathways of importance to aging, e.g., metabolism, antioxidant defense system, proteostasis, and mitochondria function. The Genomic Sciences Core uses novel next-generation sequencing that allows investigators to study the effect of age, or anti-aging manipulations, on DNA methylation, mitochondrial genome heteroplasmy, and the transcriptome of single cells. The Geroscience Redox Biology Core provides investigators with a comprehensive state-of-the-art assessment of the oxidative stress status of a cell, e.g., measures of oxidative damage and redox couples, which are important in aging as well as many major age-related diseases as well as assays of mitochondrial function. The GeroInformatics Core provides investigators assistance with data analysis, which includes both statistical support as well as analysis of large datasets. The Core also has developed number of unique software packages to help with interpretation of results and discovery of new leads relevant to aging. In addition, the Geropathology Research Resource in the Program Enhancement Core provides investigators with pathological assessments of mice using the recently developed Geropathology Grading Platform.

Insights into the Functional Role of ADTRP (Androgen-Dependent TFPI-Regulating Protein) in Health and Disease

The novel protein ADTRP, identified and described by us in 2011, is androgen-inducible and regulates the expression and activity of Tissue Factor Pathway Inhibitor, the major inhibitor of the Tissue Factor-dependent pathway of coagulation on endothelial cells. Single-nucleotide polymorphisms in ADTRP associate with coronary artery disease and myocardial infarction, and deep vein thrombosis/venous thromboembolism. Some athero-protective effects of androgen could exert through up-regulation of ADTRP expression. We discovered a critical role of ADTRP in vascular development and vessel integrity and function, manifested through Wnt signaling-dependent regulation of matrix metalloproteinase-9. ADTRP also hydrolyses fatty acid esters of hydroxy-fatty acids, which have anti-diabetic and anti-inflammatory effects and can control metabolic disorders. Here we summarize and analyze the knowledge on ADTRP and try to decipher its functions in health and disease.

A Bioinformatics Perspective on the Links Between Tetraspanin-Enriched Microdomains and Cardiovascular Pathophysiology

Background: Tetraspanins and integrins are integral membrane proteins. Tetraspanins interact with integrins to modulate the dynamics of adhesion, migration, proliferation, and signaling in the form of membrane domains called tetraspanin-enriched microdomains (TEMs). TEMs also contain other cell adhesion proteins like immunoglobulin superfamily (IgSF) proteins and claudins. Cardiovascular functions of these TEM proteins have emerged and remain to be further revealed. Objectives: The aims of this study are to explore the roles of these TEM proteins in the cardiovascular system using bioinformatics tools and databases and to highlight the TEM proteins that may functionally associate with cardiovascular physiology and pathology. Methods: For human samples, three databases-GTEx, NCBI-dbGaP, and NCBI-GEO-were used for the analyses. The dbGaP database was used for GWAS analysis to determine the association between target genes and human phenotypes. GEO is an NCBI public repository that archives genomics data. GTEx was used for the analyses of tissue-specific mRNA expression levels and eQTL. For murine samples, GeneNetwork was used to find gene-phenotype correlations and gene-gene correlations of expression levels in mice. The analysis of cardiovascular data was the focus of this study. Results: Some integrins and tetraspanins, such as ITGA8 and Cd151, are highly expressed in the human cardiovascular system. TEM components are associated with multiple cardiovascular pathophysiological events in humans. GWAS and GEO analyses showed that human Cd82 and ITGA9 are associated with blood pressure. Data from mice also suggest that various cardiovascular phenotypes are correlated with integrins and tetraspanins. For instance, Cd82 and ITGA9, again, have correlations with blood pressure in mice. Conclusion: ITGA9 is related to blood pressure in both species. KEGG analysis also linked ITGA9 to metabolism and MAPK signaling pathway. This work provides an example of using integrated bioinformatics approaches across different species to identify the connections of structurally and/or functionally related molecules to certain categories of diseases.

CSAG1 maintains the integrity of the mitotic centrosome in cells with defective p53

Centrosomes focus microtubules to promote mitotic spindle bipolarity, a critical requirement for balanced chromosome segregation. Comprehensive understanding of centrosome function and regulation requires a complete inventory of components. While many centrosome components have been identified, others yet remain undiscovered. We have used a bioinformatics approach, based on 'guilt by association' expression to identify novel mitotic components among the large group of predicted human proteins that have yet to be functionally characterized. Here, we identify chondrosarcoma-associated gene 1 protein (CSAG1) in maintaining centrosome integrity during mitosis. Depletion of CSAG1 disrupts centrosomes and leads to multipolar spindles, particularly in cells with compromised p53 function. Thus, CSAG1 may reflect a class of 'mitotic addiction' genes, whose expression is more essential in transformed cells.

Weighted Gene Correlation Network Meta-Analysis Reveals Functional Candidate Genes Associated with High- and Sub-Fertile Reproductive Performance in Beef Cattle

Improved reproductive efficiency could lead to economic benefits for the beef industry, once the intensive selection pressure has led to a decreased fertility. However, several factors limit our understanding of fertility traits, including genetic differences between populations and statistical limitations. In the present study, the RNA-sequencing data from uterine samples of high-fertile (HF) and sub-fertile (SF) animals was integrated using co-expression network meta-analysis, weighted gene correlation network analysis, identification of upstream regulators, variant calling, and network topology approaches. Using this pipeline, top hub-genes harboring fixed variants (HF × SF) were identified in differentially co-expressed gene modules (DcoExp). The functional prioritization analysis identified the genes with highest potential to be key-regulators of the DcoExp modules between HF and SF animals. Consequently, 32 functional candidate genes (10 upstream regulators and 22 top hub-genes of DcoExp modules) were identified. These genes were associated with the regulation of relevant biological processes for fertility, such as embryonic development, germ cell proliferation, and ovarian hormone regulation. Additionally, 100 candidate variants (single nucleotide polymorphisms (SNPs) and insertions and deletions (INDELs)) were identified within those genes. In the long-term, the results obtained here may help to reduce the frequency of subfertility in beef herds, reducing the associated economic losses caused by this condition.

Centralization Within Sub-Experiments Enhances the Biological Relevance of Gene Co-expression Networks: A Plant Mitochondrial Case Study

Gene co-expression networks (GCNs) can be prepared using a variety of mathematical approaches based on data sampled across diverse developmental processes, tissue types, pathologies, mutant backgrounds, and stress conditions. These networks are used to identify genes with similar expression dynamics but are prone to introducing false-positive and false-negative relationships, especially in the instance of large and heterogenous datasets. With the aim of optimizing the relevance of edges in GCNs and enhancing global biological insight, we propose a novel approach that involves a data-centering step performed simultaneously per gene and per sub-experiment, called centralization within sub-experiments (CSE). Using a gene set encoding the plant mitochondrial proteome as a case study, our results show that all CSE-based GCNs assessed had significantly more edges within the majority of the considered functional sub-networks, such as the mitochondrial electron transport chain and its complexes, than GCNs not using CSE; thus demonstrating that CSE-based GCNs are efficient at predicting canonical functions and associated pathways, here referred to as the core gene network. Furthermore, we show that correlation analyses using CSE-processed data can be used to fine-tune prediction of the function of uncharacterized genes; while its use in combination with analyses based on non-CSE data can augment conventional stress analyses with the innate connections underpinning the dynamic system being examined. Therefore, CSE is an effective alternative method to conventional batch correction approaches, particularly when dealing with large and heterogenous datasets. The method is easy to implement into a pre-existing GCN analysis pipeline and can provide enhanced biological relevance to conventional GCNs by allowing users to delineate a core gene network.

AUTHOR SUMMARY

Gene co-expression networks (GCNs) are the product of a variety of mathematical approaches that identify causal relationships in gene expression dynamics but are prone to the misdiagnoses of false-positives and false-negatives, especially in the instance of large and heterogenous datasets. In light of the burgeoning output of next-generation sequencing projects performed on a variety of species, and developmental or clinical conditions; the statistical power and complexity of these networks will undoubtedly increase, while their biological relevance will be fiercely challenged. Here, we propose a novel approach to generate a "core" GCN with enhanced biological relevance. Our method involves a data-centering step that effectively removes all primary treatment/tissue effects, which is simple to employ and can be easily implemented into pre-existing GCN analysis pipelines. The gain in biological relevance resulting from the adoption of this approach was assessed using a plant mitochondrial case study.

Optimized monoclonal antibody treatment against ELTD1 for GBM in a G55 xenograft mouse model

Glioblastoma is an aggressive brain tumour found in adults, and the therapeutic approaches available have not significantly increased patient survival. Recently, we discovered that ELTD1, an angiogenic biomarker, is highly expressed in human gliomas. Polyclonal anti-ELTD1 treatments were effective in glioma pre-clinical models, however, pAb binding is potentially promiscuous. Therefore, the aim of this study was to determine the effects of an optimized monoclonal anti-ELTD1 treatment in G55 xenograft glioma models. MRI was used to assess the effects of the treatments on animal survival, tumour volumes, perfusion rates and binding specificity. Immunohistochemistry and histology were conducted to confirm and characterize microvessel density and Notch1 levels, and to locate the molecular probes. RNA-sequencing was used to analyse the effects of the mAb treatment. Our monoclonal anti-ELTD1 treatment significantly increased animal survival, reduced tumour volumes, normalized the vasculature and showed higher binding specificity within the tumour compared with both control- and polyclonal-treated mice. Notch1 positivity staining and RNA-seq results suggested that ELTD1 has the ability to interact with and interrupt Notch1 signalling. Although little is known about ELTD1, particularly about its ligand and pathways, our data suggest that our monoclonal anti-ELTD1 antibody is a promising anti-angiogenic therapeutic in glioblastomas.

The long noncoding RNA neuroLNC regulates presynaptic activity by interacting with the neurodegeneration-associated protein TDP-43

The cellular and the molecular mechanisms by which long noncoding RNAs (lncRNAs) may regulate presynaptic function and neuronal activity are largely unexplored. Here, we established an integrated screening strategy to discover lncRNAs implicated in neurotransmitter and synaptic vesicle release. With this approach, we identified neuroLNC, a neuron-specific nuclear lncRNA conserved from rodents to humans. NeuroLNC is tuned by synaptic activity and influences several other essential aspects of neuronal development including calcium influx, neuritogenesis, and neuronal migration in vivo. We defined the molecular interactors of neuroLNC in detail using chromatin isolation by RNA purification, RNA interactome analysis, and protein mass spectrometry. We found that the effects of neuroLNC on synaptic vesicle release require interaction with the RNA-binding protein TDP-43 (TAR DNA binding protein-43) and the selective stabilization of mRNAs encoding for presynaptic proteins. These results provide the first proof of an lncRNA that orchestrates neuronal excitability by influencing presynaptic function.

Characterization of cxorf21 Provides Molecular Insight Into Female-Bias Immune Response in SLE Pathogenesis

Background: Ninety percent of systemic lupus erythematosus (SLE) patients are women. X chromosome-dosage increases susceptibility to SLE and primary Sjögren's syndrome (pSS). Chromosome X open reading frame 21 (CXorf21) escapes X-inactivation and is an SLE risk gene of previously unknown function. We undertook the present study to delineate the function of CXorf21 in the immune system as well as investigate a potential role in the sex bias of SLE and pSS. Methods: Western blot protein analysis, qPCR, BioPlex cytokine immunoassay, pHrodo™ assays, as well as in vitro CRISPR-Cas9 knockdown experiments were employed to delineate the role of CXorf21 in relevant immunocytes. Results: Expressed in monocytes and B cells, CXorf21 basal Mrna, and protein expression levels are elevated in female primary monocytes, B cells, and EBV-transformed B cells compared to male cells. We also found CXorf21 mRNA and protein expression is higher in both male and female cells from SLE patients compared to control subjects. TLR7 ligation increased CXorf21 protein expression and CXorf21 knockdown abrogated TLR7-driven increased IFNA1 mRNA expression, and reduced secretion of both TNF-alpha and IL-6 in healthy female monocytes. Similarly, we found increased pH in the lysosomes of CXorf21-deficient female monocytes. Conclusion: CXorf21 is more highly expressed in female compared to male cells and is involved in a sexually dimorphic response to TLR7 activation. In addition, CXorf21 expression regulates lysosomal pH in a sexually dimorphic manner. Thus, sexually dimorphic expression of CXorf21 skews cellular immune responses in manner consistent with expected properties of a mediator of the X chromosome dose risk in SLE and pSS.

Integrating sequence and gene expression information predicts genome-wide DNA-binding proteins and suggests a cooperative mechanism

DNA-binding proteins (DBPs) perform diverse biological functions ranging from transcription to pathogen sensing. Machine learning methods can not only identify DBPs de novo but also provide insights into their DNA-recognition dynamics. However, it remains unclear whether available methods that can accurately predict DNA-binding sites in known DBPs can also identify novel DBPs. Moreover, sequence information is blind to the cellular- and disease-specific contexts of DBP activities, whereas the under-utilized knowledge from public gene expression data offers great promise. To address these issues, we have developed novel methods for predicting DBPs by integrating sequence and gene expression-derived features and applied them to explore human, mouse and Arabidopsis proteomes. While our sequence-based models outperformed the gene expression-based ones, some proteins with weaker DBP-like sequence features were correctly predicted by gene expression-based features, suggesting that these proteins acquire a tangible DBP functionality in a conducive gene expression environment. Analysis of motif enrichment among the co-expressed genes of top 100 candidates DBPs from hitherto unannotated genes provides further avenues to explore their functional associations.

Disease classification: from phenotypic similarity to integrative genomics and beyond

A fundamental challenge of modern biomedical research is understanding how diseases that are similar on the phenotypic level are similar on the molecular level. Integration of various genomic data sets with the traditionally used phenotypic disease similarity revealed novel genetic and molecular mechanisms and blurred the distinction between monogenic (Mendelian) and complex diseases. Network-based medicine has emerged as a complementary approach for identifying disease-causing genes, genetic mediators, disruptions in the underlying cellular functions and for drug repositioning. The recent development of machine and deep learning methods allow for leveraging real-life information about diseases to refine genetic and phenotypic disease relationships. This review describes the historical development and recent methodological advancements for studying disease classification (nosology).

Machine learning technology in the application of genome analysis: A systematic review

Machine learning (ML) is a powerful technique to tackle many problems in data mining and predictive analytics. We believe that ML will be of considerable potentials in the field of bioinformatics since the high-throughput technology is producing ever increasing biological data. In this review, we summarized major ML algorithms and conditions that must be paid attention to when applying these algorithms to genomic problems in details and we provided a list of examples from different perspectives and data analysis challenges at present.

Targeting ELTD1, an angiogenesis marker for glioblastoma (GBM), also affects VEGFR2: molecular-targeted MRI assessment

Glioblastomas (GBM) are deadly brain tumors that currently do not have long-term patient treatments available. GBM overexpress the angiogenesis factor VEGF and its receptor VEGFR2. ETLD1 (epidermal growth factor, latrophilin and seven transmembrane domain-containing protein 1), a G-protein coupled receptor (GPCR) protein, we discovered as a biomarker for high-grade gliomas, is also a novel regulator of angiogenesis. Since it was established that VEGF regulates ELTD1, we wanted to establish if VEGFR2 is also associated with ELTD1, by targeted antibody inhibition. G55 glioma-bearing mice were treated with either anti-ELTD1 or anti-VEGFR2 antibodies. With the use of MRI molecular imaging probes, we were able to detect in vivo levels of either ELTD1 (anti-ELTD1 probe) or VEGFR2 (anti-VEGFR2 probe). Protein expressions were obtained for ELTD1, VEGF or VEGFR2 via immunohistochemistry (IHC). VEGFR2 levels were significantly decreased (P < 0.05) with anti-ELTD1 antibody treatment, and ELTD1 levels were significantly decreased (P < 0.05) with anti-VEGFR2 antibody treatment, both compared to untreated tumors. IHC from mouse tumor tissues established that VEGFR2 and ELTD1 were co-localized. The mouse anti-ELTD1 antibody treatment study indicated that anti-VEGFR2 antibody treatment does not significantly increase survival, decrease tumor volumes, or alter tumor perfusion (measured as relative cerebral blood flow or rCBF). Conversely, anti-ELTD1 antibody treatment was able to significantly increase animal survival (P < 0.01), decrease tumor volumes (P < 0.05), and reduce change in rCBF (P < 0.001), when compared to untreated or IgG-treated tumor bearing mice. Anti-ELTD1 antibody therapy could be beneficial in targeting ELTD1, as well as simultaneously affecting VEGFR2, as a possible GBM treatment.

Comparative Analysis of Developmental Transcriptome Maps of Arabidopsis thaliana and Solanum lycopersicum

The knowledge of gene functions in model organisms is the starting point for the analysis of gene function in non-model species, including economically important ones. Usually, the assignment of gene functions is based on sequence similarity. In plants, due to a highly intricate gene landscape, this approach has some limitations. It is often impossible to directly match gene sets from one plant species to another species based only on their sequences. Thus, it is necessary to use additional information to identify functionally similar genes. Expression patterns have great potential to serve as a source of such information. An important prerequisite for the comparative analysis of transcriptomes is the existence of high-resolution expression maps consisting of comparable samples. Here, we present a transcriptome atlas of tomato (Solanum lycopersicum) consisting of 30 samples of different organs and developmental stages. The samples were selected in a way that allowed for side-by-side comparison with the Arabidopsis thaliana transcriptome map. Newly obtained data are integrated in the TraVA database and are available online, together with tools for their analysis. In this paper, we demonstrate the potential of comparing transcriptome maps for inferring shifts in the expression of paralogous genes.

Interpretation of an individual functional genomics experiment guided by massive public data

A key unmet challenge in interpreting omics experiments is inferring biological meaning in the context of public functional genomics data. We developed a computational framework, Your Evidence Tailored Integration (YETI; http://yeti.princeton.edu/ ), which creates specialized functional interaction maps from large public datasets relevant to an individual omics experiment. Using this tailored integration, we predicted and experimentally confirmed an unexpected divergence in viral replication after seasonal or pandemic human influenza virus infection.

Tetraspanin-enriched microdomains regulate digitation junctions

Tetraspanins co-emerged with multi-cellular organisms during evolution and are typically localized at the cell–cell interface, [corrected] and form tetraspanin-enriched microdomains (TEMs) by associating with each other and other membrane molecules. Tetraspanins affect various biological functions, but how tetraspanins engage in multi-faceted functions at the cellular level is largely unknown. When cells interact, the membrane microextrusions at the cell-cell interfaces form dynamic, digit-like structures between cells, which we term digitation junctions (DJs). We found that (1) tetraspanins CD9, CD81, and CD82 and (2) TEM-associated molecules integrin α3β1, CD44, EWI2/PGRL, and PI-4P are present in DJs of epithelial, endothelial, and cancer cells. Tetraspanins and their associated molecules also regulate the formation and development of DJs. Moreover, (1) actin cytoskeleton, RhoA, and actomyosin activities and (2) growth factor receptor-Src-MAP kinase signaling, but not PI-3 kinase, regulate DJs. Finally, we showed that DJs consist of various forms in different cells. Thus, DJs are common, interactive structures between cells, and likely affect cell adhesion, migration, and communication. TEMs probably modulate various cell functions through DJs. Our findings highlight that DJ morphogenesis reflects the transition between cell-matrix adhesion and cell-cell adhesion and involves both cell-cell and cell-matrix adhesion molecules.

Consensus Coexpression Network Analysis Identifies Key Regulators of Flower and Fruit Development in Wild Strawberry

The diploid strawberry, Fragaria vesca, is a developing model system for the economically important Rosaceae family. Strawberry fleshy fruit develops from the floral receptacle and its ripening is nonclimacteric. The external seed configuration of strawberry fruit facilitates the study of seed-to-fruit cross tissue communication, particularly phytohormone biosynthesis and transport. To investigate strawberry fruit development, we previously generated spatial and temporal transcriptome data profiling F. vesca flower and fruit development pre- and postfertilization. In this study, we combined 46 of our existing RNA-seq libraries to generate coexpression networks using the Weighted Gene Co-Expression Network Analysis package in R. We then applied a post-hoc consensus clustering approach and used bootstrapping to demonstrate consensus clustering's ability to produce robust and reproducible clusters. Further, we experimentally tested hypotheses based on the networks, including increased iron transport from the receptacle to the seed postfertilization and characterized a F. vesca floral mutant and its candidate gene. To increase their utility, the networks are presented in a web interface (www.fv.rosaceaefruits.org) for easy exploration and identification of coexpressed genes. Together, the work reported here illustrates ways to generate robust networks optimized for the mining of large transcriptome data sets, thereby providing a useful resource for hypothesis generation and experimental design in strawberry and related Rosaceae fruit crops.

APache Is an AP2-Interacting Protein Involved in Synaptic Vesicle Trafficking and Neuronal Development

Synaptic transmission is critically dependent on synaptic vesicle (SV) recycling. Although the precise mechanisms of SV retrieval are still debated, it is widely accepted that a fundamental role is played by clathrin-mediated endocytosis, a form of endocytosis that capitalizes on the clathrin/adaptor protein complex 2 (AP2) coat and several accessory factors. Here, we show that the previously uncharacterized protein KIAA1107, predicted by bioinformatics analysis to be involved in the SV cycle, is an AP2-interacting clathrin-endocytosis protein (APache). We found that APache is highly enriched in the CNS and is associated with clathrin-coated vesicles via interaction with AP2. APache-silenced neurons exhibit a severe impairment of maturation at early developmental stages, reduced SV density, enlarged endosome-like structures, and defects in synaptic transmission, consistent with an impaired clathrin/AP2-mediated SV recycling. Our data implicate APache as an actor in the complex regulation of SV trafficking, neuronal development, and synaptic plasticity.

Predictive bioinformatics identifies novel regulators of proliferation in a cancer stem cell model

The cancer stem cell model postulates that tumors are hierarchically organized with a minor population, the cancer stem cells, exhibiting unlimited proliferative potential. These cells give rise to the bulk of tumor cells, which retain a limited ability to divide. Without successful targeting of cancer stem cells, tumor reemergence after therapy is likely. However, identifying target pathways essential for cancer stem cell proliferation has been challenging. Here, using a transcriptional network analysis termed GAMMA, we identified 50 genes whose correlation patterns suggested involvement in cancer stem cell division. Using RNAi depletion, we found that 21 of these target genes showed preferential growth inhibition in a breast cancer stem cell model. More detailed initial analysis of 6 of these genes revealed 4 with clear roles in the fidelity of chromosome segregation. This study reveals the strong predictive potential of transcriptional network analysis in increasing the efficiency of successful identification of novel proliferation dependencies for cancer stem cells.

GTSE1 regulates spindle microtubule dynamics to control Aurora B kinase and Kif4A chromokinesin on chromosome arms

In mitosis, the dynamic assembly and disassembly of microtubules are critical for normal chromosome movement and segregation. Microtubule turnover varies among different mitotic spindle microtubules, dictated by their spatial distribution within the spindle. How turnover among the various classes of spindle microtubules is differentially regulated and the resulting significance of differential turnover for chromosome movement remains a mystery. As a new tactic, we used global microarray meta-analysis (GAMMA), a bioinformatic method, to identify novel regulators of mitosis, and in this study, we describe G2- and S phase-expressed protein 1 (GTSE1). GTSE1 is expressed exclusively in late G2 and M phase. From nuclear envelope breakdown until anaphase onset, GTSE1 binds preferentially to the most stable mitotic spindle microtubules and promotes their turnover. Cells depleted of GTSE1 show defects in chromosome alignment at the metaphase plate and in spindle pole integrity. These defects are coupled with an increase in the proportion of stable mitotic spindle microtubules. A consequence of this reduced microtubule turnover is diminished recruitment and activity of Aurora B kinase on chromosome arms. This decrease in Aurora B results in diminished binding of the chromokinesin Kif4A to chromosome arms.

ELTD1, an effective anti-angiogenic target for gliomas: preclinical assessment in mouse GL261 and human G55 xenograft glioma models

Background

Despite current therapies, glioblastoma is a devastating cancer, and validation of effective biomarkers for it will enable better diagnosis and therapeutic intervention for this disease. We recently discovered a new biomarker for high-grade gliomas, ELTD1 (epidermal growth factor, latrophilin, and 7 transmembrane domain-containing protein 1 on chromosome 1) via bioinformatics, and validated that ELTD1 protein levels are significantly higher in human and rodent gliomas. The focus of this study was to assess the effect on tumor growth of an antibody against ELTD1 in orthotopic, GL261, and G55 xenograft glioma models.

Methods

The effect of anti-ELTD1 antibody therapy was assessed by animal survival, MRI measured tumor volumes, MR angiography, MR perfusion imaging, and immunohistochemistry (IHC) characterization of microvessel density in mouse glioma models. Comparative treatments included anti-vascular endothelial growth factor (VEGF) and anti-c-Met antibody therapies, compared with untreated controls.

Results

Tumor volume and survival data in this study show that antibodies against ELTD1 inhibit glioma growth just as effectively or even more so compared with other therapeutic targets studied, including anti-VEGF antibody therapy. Untreated GL261 or G55 tumors were found to have significantly higher ELTD1 levels (IHC) compared with contralateral normal brain. The anti-angiogenic effect of ELTD1 antibody therapy was observed in assessment of microvessel density, as well as from MR angiography and perfusion measurements, which indicated that anti-ELTD1 antibody therapy significantly decreased vascularization compared with untreated controls.

Conclusions

Either as a single therapy or in conjunction with other therapeutic approaches, anti-ELTD1 antibodies could be a valuable new clinical anti-angiogenic therapeutic for high-grade gliomas.

The Development of Translational Biomarkers as a Tool for Improving the Understanding, Diagnosis and Treatment of Chronic Neuropathic Pain

Chronic neuropathic pain (CNP) is one of the most significant unmet clinical needs in modern medicine. Alongside the lack of effective treatments, there is a great deficit in the availability of objective diagnostic methods to reliably facilitate an accurate diagnosis. We therefore aimed to determine the feasibility of a simple diagnostic test by analysing differentially expressed genes in the blood of patients diagnosed with CNP of the lower back and compared to healthy human controls. Refinement of microarray expression data was performed using correlation analysis with 3900 human 2-colour microarray experiments. Selected genes were analysed in the dorsal horn of Sprague-Dawley rats after L5 spinal nerve ligation (SNL), using qRT-PCR and ddPCR, to determine possible associations with pathophysiological mechanisms underpinning CNP and whether they represent translational biomarkers of CNP. We found that of the 15 potential biomarkers identified, tissue inhibitor of matrix metalloproteinase-1 (TIMP1) gene expression was upregulated in chronic neuropathic lower back pain (CNBP) (p = 0.0049) which positively correlated (R = 0.68, p = ≤0.05) with increased plasma TIMP1 levels in this group (p = 0.0433). Moreover, plasma TIMP1 was also significantly upregulated in CNBP than chronic inflammatory lower back pain (p = 0.0272). In the SNL model, upregulation of the Timp1 gene was also observed (p = 0.0058) alongside a strong trend for the upregulation of melanocortin 1 receptor (p = 0.0847). Our data therefore highlights several genes that warrant further investigation, and of these, TIMP1 shows the greatest potential as an accessible and translational CNP biomarker.

The Role of Neutrophil Proteins on the Amyloid Beta-RAGE Axis

We previously showed an elevated expression of the neutrophil protein, cationic antimicrobial protein of 37kDa (CAP37), in brains of patients with Alzheimer’s disease (AD), suggesting that CAP37 could be involved in AD pathogenesis. The first step in determining how CAP37 might contribute to AD pathogenesis was to identify the receptor through which it induces cell responses. To identify a putative receptor, we performed GAMMA analysis to determine genes that positively correlated with CAP37 in terms of expression. Positive correlations with ligands for the receptor for advanced glycation end products (RAGE) were observed. Additionally, CAP37 expression positively correlated with two other neutrophil proteins, neutrophil elastase and cathepsin G. Enzyme-linked immunosorbent assays (ELISAs) demonstrated an interaction between CAP37, neutrophil elastase, and cathepsin G with RAGE. Amyloid beta 1–42 (Aβ_1–42), a known RAGE ligand, accumulates in AD brains and interacts with RAGE, contributing to Aβ_1–42 neurotoxicity. We questioned whether the binding of CAP37, neutrophil elastase and/or cathepsin G to RAGE could interfere with Aβ_1–42 binding to RAGE. Using ELISAs, we determined that CAP37 and neutrophil elastase inhibited binding of Aβ_1–42 to RAGE, and this effect was reversed by protease inhibitors in the case of neutrophil elastase. Since neutrophil elastase and cathepsin G have enzymatic activity, mass spectrometry was performed to determine the proteolytic activity of all three neutrophil proteins on Aβ_1–42. All three neutrophil proteins bound to Aβ_1–42 with different affinities and cleaved Aβ_1–42 with different kinetics and substrate specificities. We posit that these neutrophil proteins could modulate neurotoxicity in AD by cleaving Aβ_1–42 and influencing the Aβ1–42 –RAGE interaction. Further studies will be required to determine the biological significance of these effects and their relevance in neurodegenerative diseases such as AD. Our findings identify a novel area of study that underscores the importance of neutrophils and neutrophil proteins in neuroinflammatory diseases such as AD.

LncRNA ontology: inferring lncRNA functions based on chromatin states and expression patterns

Accumulating evidences suggest that long non-coding RNAs (lncRNAs) perform important functions. Genome-wide chromatin-states area rich source of information about cellular state, yielding insights beyond what is typically obtained by transcriptome profiling. We propose an integrative method for genome-wide functional predictions of lncRNAs by combining chromatin states data with gene expression patterns. We first validated the method using protein-coding genes with known function annotations. Our validation results indicated that our integrative method performs better than co-expression analysis, and is accurate across different conditions. Next, by applying the integrative model genome-wide, we predicted the probable functions for more than 97% of human lncRNAs. The putative functions inferred by our method match with previously annotated by the targets of lncRNAs. Moreover, the linkage from the cellular processes influenced by cancer-associated lncRNAs to the cancer hallmarks provided a “lncRNA point-of-view” on tumor biology. Our approach provides a functional annotation of the lncRNAs, which we developed into a web-based application, LncRNA Ontology, to provide visualization, analysis, and downloading of lncRNA putative functions.

Long Non-Coding RNAs (lncRNAs) of Sea Cucumber: Large-Scale Prediction, Expression Profiling, Non-Coding Network Construction, and lncRNA-microRNA-Gene Interaction Analysis of lncRNAs in Apostichopus japonicus and Holothuria glaberrima During LPS Challenge and Radial Organ Complex Regeneration

Long non-coding RNA (lncRNA) structurally resembles mRNA but cannot be translated into protein. Although the systematic identification and characterization of lncRNAs have been increasingly reported in model species, information concerning non-model species is still lacking. Here, we report the first systematic identification and characterization of lncRNAs in two sea cucumber species: (1) Apostichopus japonicus during lipopolysaccharide (LPS) challenge and in heathy tissues and (2) Holothuria glaberrima during radial organ complex regeneration, using RNA-seq datasets and bioinformatics analysis. We identified A. japonicus and H. glaberrima lncRNAs that were differentially expressed during LPS challenge and radial organ complex regeneration, respectively. Notably, the predicted lncRNA-microRNA-gene trinities revealed that, in addition to targeting protein-coding transcripts, miRNAs might also target lncRNAs, thereby participating in a potential novel layer of regulatory interactions among non-coding RNA classes in echinoderms. Furthermore, the constructed coding-non-coding network implied the potential involvement of lncRNA-gene interactions during the regulation of several important genes (e.g., Toll-like receptor 1 [TLR1] and transglutaminase-1 [TGM1]) in response to LPS challenge and radial organ complex regeneration in sea cucumbers. Overall, this pioneer systematic identification, annotation, and characterization of lncRNAs in echinoderm pave the way for similar studies and future genetic, genomic, and evolutionary research in non-model species.

IGF-1 deficiency in a critical period early in life influences the vascular aging phenotype in mice by altering miRNA-mediated post-transcriptional gene regulation: implications for the developmental origins of health and disease hypothesis

Epidemiological findings support the concept of Developmental Origins of Health and Disease, suggesting that early-life hormonal influences during a sensitive period of development have a fundamental impact on vascular health later in life. The endocrine changes that occur during development are highly conserved across mammalian species and include dramatic increases in circulating IGF-1 levels during adolescence. The present study was designed to characterize the effect of developmental IGF-1 deficiency on the vascular aging phenotype. To achieve that goal, early-onset endocrine IGF-1 deficiency was induced in mice by knockdown of IGF-1 in the liver using Cre-lox technology (Igf1 ^f/f mice crossed with mice expressing albumin-driven Cre recombinase). This model exhibits low-circulating IGF-1 levels during the peripubertal phase of development, which is critical for the biology of aging. Due to the emergence of miRNAs as important regulators of the vascular aging phenotype, the effect of early-life IGF-1 deficiency on miRNA expression profile in the aorta was examined in animals at 27 months of age. We found that developmental IGF-1 deficiency elicits persisting late-life changes in miRNA expression in the vasculature, which significantly differed from those in mice with adult-onset IGF-1 deficiency (TBG-Cre-AAV8-mediated knockdown of IGF-1 at 5 month of age in Igf1 ^f/f mice). Using a novel computational approach, we identified miRNA target genes that are co-expressed with IGF-1 and associate with aging and vascular pathophysiology. We found that among the predicted targets, the expression of multiple extracellular matrix-related genes, including collagen-encoding genes, were downregulated in mice with developmental IGF-1 deficiency. Collectively, IGF-1 deficiency during a critical period during early in life results in persistent changes in post-transcriptional miRNA-mediated control of genes critical targets for vascular health, which likely contribute to the deleterious late-life cardiovascular effects known to occur with developmental IGF-1 deficiency.

Upregulation of PSMD10 caused by the JMJD2A histone demethylase

PSMD10, also known as gankyrin, is associated with the proteasome and has been shown to be an oncoprotein in the liver. Here, we report that PSMD10 expression is stimulated by the histone demethylase JMJD2A/KDM4A and its interaction partner, the ETV1 transcription factor, in LNCaP prostate cancer cells. Global analysis of expression patterns revealed that PSMD10 mRNA levels are positively correlated with those of both JMJD2A and ETV1. In human prostate tumors, PSMD10 is highly overexpressed at the protein level and correlates with JMJD2A overexpression; further, PSMD10 expression is enhanced in the prostates of transgenic JMJD2A mice. Moreover, PSMD10 is particularly overexpressed in high Gleason score prostate tumors. Downregulation of PSMD10 in LNCaP prostate cancer cells impaired their growth, indicating that PSMD10 may exert a pro-oncogenic function in the prostate. Lastly, we observed that PSMD10 expression is correlated to YAP1, a component of the Hippo signaling pathway and whose gene promoter is regulated by JMJD2A, and that PSMD10 can cooperate with YAP1 in stimulating LNCaP cell growth. Altogether, these data indicate that PSMD10 is a novel downstream effector of JMJD2A and suggest that inhibition of the JMJD2A histone demethylase by small molecule drugs may be effective to curtail the oncogenic activity of PSMD10 in various PSMD10-overexpressing tumors.

Long noncoding RNAs expressed in human hepatic stellate cells form networks with extracellular matrix proteins

BACKGROUND

Hepatic fibrosis is the underlying cause of cirrhosis and liver failure in nearly every form of chronic liver disease, and hepatic stellate cells (HSCs) are the primary cell type responsible for fibrosis. Long noncoding RNAs (lncRNAs) are increasingly recognized as regulators of development and disease; however, little is known about their expression in human HSCs and their function in hepatic fibrosis.

METHODS

We performed RNA sequencing and ab initio assembly of RNA transcripts to define the lncRNAs expressed in human HSC myofibroblasts. We analyzed chromatin immunoprecipitation data and expression data to identify lncRNAs that were regulated by transforming growth factor beta (TGF-β) signaling, associated with super-enhancers and restricted in expression to HSCs compared with 43 human tissues and cell types. Co-expression network analyses were performed to discover functional modules of lncRNAs, and principle component analysis and K-mean clustering were used to compare lncRNA expression in HSCs with other myofibroblast cell types.

RESULTS

We identified over 3600 lncRNAs that are expressed in human HSC myofibroblasts. Many are regulated by TGF-β, a major fibrotic signal, and form networks with genes encoding key components of the extracellular matrix (ECM), which is the substrate of the fibrotic scar. The lncRNAs directly regulated by TGF-β signaling are also enriched at super-enhancers. More than 400 of the lncRNAs identified in HSCs are uniquely expressed in HSCs compared with 43 other human tissues and cell types and HSC myofibroblasts demonstrate different patterns of lncRNA expression compared with myofibroblasts originating from other tissues. Co-expression analyses identified a subset of lncRNAs that are tightly linked to collagen genes and numerous proteins that regulate the ECM during formation of the fibrotic scar. Finally, we identified lncRNAs that are induced during progression of human liver disease.

CONCLUSIONS

lncRNAs are likely key contributors to the formation and progression of fibrosis in human liver disease.

Coding-noncoding gene expression in intrahepatic cholangiocarcinoma

Recent studies have shown that long noncoding RNAs (lncRNAs) play crucial roles in human cancers. However, the function of lncRNAs and their downstream mechanisms are largely unknown in the molecular pathogenesis of intrahepatic cholangiocarcinoma (ICC). In the present study, we performed transcriptomic profiling of ICC and paired adjacent noncancerous tissues (N) by using lncRNA and messenger RNA (mRNA) microarrays. Quantitative real-time polymerase chain reaction was used to validate the microarray results. We tested for correlations between the expression levels of lncRNAs and target genes. Clinicopathologic characteristics and overall survival were compared using the t test and the Kaplan-Meier method, respectively. A total of 2773 lncRNAs were significantly upregulated in ICC tissues compared with the noncancerous tissues, whereas 2392 lncRNAs were downregulated. Bioinformatic analysis indicated that most of the genes were involved in carcinogenesis, hepatic system diseases, and signal transductions. Positive correlations were found between 4 lncRNA-mRNA pairs (RNA43085 and SULF1, RNA47504 and KDM8, RNA58630 and PCSK6, and RNA40057 and CYP2D6). When the clinicopathologic characteristics were accounted for, the cumulative overall survival rate was found to be associated with low expression levels of CYP2D6 (P = 0.005) and PCSK6 (P = 0.038). Patients with high expression levels of CYP2D6 and RNA40057 had a better prognosis (P = 0.014). Our results suggested that the lncRNA expression profiling in ICC tissues is profoundly different from that in noncancerous tissues. Thus, lncRNA may be a potential diagnostic and prognostic biomarker for ICC. Furthermore, the combined assessment of lncRNA and mRNA expressions might predict the survival of patients with ICC.

Histone demethylase JMJD2A drives prostate tumorigenesis through transcription factor ETV1

Histone demethylase upregulation has been observed in human cancers, yet it is unknown whether this is a bystander event or a driver of tumorigenesis. We found that overexpression of lysine-specific demethylase 4A (KDM4A, also known as JMJD2A) was positively correlated with Gleason score and metastasis in human prostate tumors. Overexpression of JMJD2A resulted in the development of prostatic intraepithelial neoplasia in mice, demonstrating that JMJD2A can initiate prostate cancer development. Moreover, combined overexpression of JMJD2A and the ETS transcription factor ETV1, a JMJD2A-binding protein, resulted in prostate carcinoma formation in mice haplodeficient for the phosphatase and tensin homolog (Pten) tumor-suppressor gene. Additionally, JMJD2A cooperated with ETV1 to increase expression of yes associated protein 1 (YAP1), a Hippo pathway component that itself was associated with prostate tumor aggressiveness. ETV1 facilitated the recruitment of JMJD2A to the YAP1 promoter, leading to changes in histone lysine methylation in a human prostate cancer cell line. Further, YAP1 expression largely rescued the growth inhibitory effects of JMJD2A depletion in prostate cancer cells, indicating that YAP1 is a downstream effector of JMJD2A. Taken together, these data reveal a JMJD2A/ETV1/YAP1 axis that promotes prostate cancer initiation and that may be a suitable target for therapeutic inhibition.

Analysis of TMEFF2 allografts and transgenic mouse models reveals roles in prostate regeneration and cancer

BACKGROUND

Previous results from our lab indicate a tumor suppressor role for the transmembrane protein with epidermal growth factor and two follistatin motifs 2 (TMEFF2) in prostate cancer (PCa). Here, we further characterize this role and uncover new functions for TMEFF2 in cancer and adult prostate regeneration.

METHODS

The role of TMEFF2 was examined in PCa cells using Matrigel(TM) cultures and allograft models of PCa cells. In addition, we developed a transgenic mouse model that expresses TMEFF2 from a prostate specific promoter. Anatomical, histological, and metabolic characterizations of the transgenic mouse prostate were conducted. The effect of TMEFF2 in prostate regeneration was studied by analyzing branching morphogenesis in the TMEFF2-expressing mouse lobes and alterations in branching morphogenesis were correlated with the metabolomic profiles of the mouse lobes. The role of TMEFF2 in prostate tumorigenesis in whole animals was investigated by crossing the TMEFF2 transgenic mice with the TRAMP mouse model of PCa and analyzing the histopathological changes in the progeny.

RESULTS

Ectopic expression of TMEFF2 impairs growth of PCa cells in Matrigel or allograft models. Surprisingly, while TMEFF2 expression in the TRAMP mouse did not have a significant effect on the glandular prostate epithelial lesions, the double TRAMP/TMEFF2 transgenic mice displayed an increased incidence of neuroendocrine type tumors. In addition, TMEFF2 promoted increased branching specifically in the dorsal lobe of the prostate suggesting a potential role in developmental processes. These results correlated with data indicating an alteration in the metabolic profile of the dorsal lobe of the transgenic TMEFF2 mice.

CONCLUSIONS

Collectively, our results confirm the tumor suppressor role of TMEFF2 and suggest that ectopic expression of TMEFF2 in mouse prostate leads to additional lobe-specific effects in prostate regeneration and tumorigenesis. This points to a complex and multifunctional role for TMEFF2 during PCa progression.

Ethnicity-specific epigenetic variation in naïve CD4+ T cells and the susceptibility to autoimmunity

Background

Genetic and epigenetic variability contributes to the susceptibility and pathogenesis of autoimmune diseases. T cells play an important role in several autoimmune conditions, including lupus, which is more common and more severe in people of African descent. To investigate inherent epigenetic differences in T cells between ethnicities, we characterized genome-wide DNA methylation patterns in naïve CD4+ T cells in healthy African-Americans and European-Americans, and then confirmed our findings in lupus patients.

Results

Impressive ethnicity-specific clustering of DNA methylation profiling in naïve CD4+ T cells was revealed. Hypomethylated loci in healthy African-Americans were significantly enriched in pro-apoptotic and pro-inflammatory genes. We also found hypomethylated genes in African-Americans to be disproportionately related to autoimmune diseases including lupus. We then confirmed that these genes, such as IL32, CD226, CDKN1A, and PTPRN2 were similarly hypomethylated in lupus patients of African-American compared to European-American descent. Using patch DNA methylation and luciferase reporter constructs, we showed that methylation of the IL32 promoter region reduces gene expression in vitro. Importantly, bisulfite DNA sequencing demonstrated that cis-acting genetic variants within and directly disrupting CpG sites account for some ethnicity-specific variability in DNA methylation.

Conclusion

Ethnicity-specific inherited epigenetic susceptibility loci in CD4+ T cells provide clues to explain differences in the susceptibility to autoimmunity and possibly other T cell-related diseases between populations.

Electronic supplementary material

The online version of this article (doi:10.1186/s13072-015-0037-1) contains supplementary material, which is available to authorized users.

A comparison of human and mouse gene co-expression networks reveals conservation and divergence at the tissue, pathway and disease levels

Background

A deeper understanding of differences and similarities in transcriptional regulation between species can uncover important information about gene functions and the role of genes in disease. Deciphering such patterns between mice and humans is especially important since mice play an essential role in biomedical research.

Results

Here, in order to characterize evolutionary changes between humans and mice, we compared gene co-expression maps to evaluate the conservation of co-expression. We show that the conservation of co-expression connectivity of homologous genes is negatively correlated with molecular evolution rates, as expected. Then we investigated evolutionary aspects of gene sets related to functions, tissues, pathways and diseases. Genes expressed in the testis, eye and skin, and those associated with regulation of transcription, olfaction, PI3K signalling, response to virus and bacteria were more divergent between mice and humans in terms of co-expression connectivity. Surprisingly, a deeper investigation of the PI3K signalling cascade revealed that its divergence is caused by the most crucial genes of this pathway, such as mTOR and AKT2. On the other hand, our analysis revealed that genes expressed in the brain and in the bone, and those associated with cell adhesion, cell cycle, DNA replication and DNA repair are most strongly conserved in terms of co-expression network connectivity as well as having a lower rate of duplication events. Genes involved in lipid metabolism and genes specific to blood showed a signature of increased co-expression connectivity in the mouse. In terms of diseases, co-expression connectivity of genes related to metabolic disorders is the most strongly conserved between mice and humans and tumor-related genes the most divergent.

Conclusions

This work contributes to discerning evolutionary patterns between mice and humans in terms of gene interactions. Conservation of co-expression is a powerful approach to identify gene targets and processes with potential similarity and divergence between mice and humans, which has implications for drug testing and other studies employing the mouse as a model organism.

Electronic supplementary material

The online version of this article (doi:10.1186/s12862-015-0534-7) contains supplementary material, which is available to authorized users.

Genetic Variation in the Platelet Endothelial Aggregation Receptor 1 Gene Results in Endothelial Dysfunction

Platelet Endothelial Aggregation Receptor 1 (PEAR1) is a newly identified membrane protein reported to be involved in multiple vascular and thrombotic processes. While most studies to date have focused on the effects of this receptor in platelets, PEAR1 is located in multiple tissues including the endothelium, where it is most highly expressed. Our first objective was to evaluate the role of PEAR1 in endothelial function by examining flow-mediated dilation of the brachial artery in 641 participants from the Heredity and Phenotype Intervention Heart Study. Our second objective was to further define the impact of PEAR1 on cardiovascular disease computationally through meta-analysis of 75,000 microarrays, yielding insights regarding PEAR1 function, and predictions of phenotypes and diseases affected by PEAR1 dysregulation. Based on the results of this meta-analysis we examined whether genetic variation in PEAR1 influences endothelial function using an ex vivo assay of endothelial cell migration. We observed a significant association between rs12041331 and flow-mediated dilation in participants of the Heredity and Phenotype Intervention Heart Study (P = 0.02). Meta-analysis results revealed that PEAR1 expression is highly correlated with several genes (e.g. ANG2, ACVRL1, ENG) and phenotypes (e.g. endothelial cell migration, angiogenesis) that are integral to endothelial function. Functional validation of these results revealed that PEAR1 rs12041331 is significantly associated with endothelial migration (P = 0.04). Our results suggest for the first time that genetic variation of PEAR1 is a significant determinant of endothelial function through pathways implicated in cardiovascular disease.

IL-18 acts in synergy with IL-7 to promote ex vivo expansion of T lymphoid progenitor cells

Although IL-18 has not previously been shown to promote T lymphopoiesis, results obtained via a novel data mining algorithm (global microarray meta-analysis) led us to explore a predicted role for this cytokine in T cell development. IL-18 is a member of the IL-1 cytokine family that has been extensively characterized as a mediator of inflammatory immune responses. To assess a potential role for IL-18 in T cell development, we sort-purified mouse bone marrow-derived common lymphoid progenitor cells, early thymic progenitors (ETPs), and double-negative 2 thymocytes and cultured these populations on OP9-Delta-like 4 stromal layers in the presence or absence of IL-18 and/or IL-7. After 1 wk of culture, IL-18 promoted proliferation and accelerated differentiation of ETPs to the double-negative 3 stage, similar in efficiency to IL-7. IL-18 showed synergy with IL-7 and enhanced proliferation of both the thymus-derived progenitor cells and the bone marrow-derived common lymphoid progenitor cells. The synergistic effect on the ETP population was further characterized and found to correlate with increased surface expression of c-Kit and IL-7 receptors on the IL-18-treated cells. In summary, we successfully validated the global microarray meta-analysis prediction that IL-18 affects T lymphopoiesis and demonstrated that IL-18 can positively impact bone marrow lymphopoiesis and T cell development, presumably via interaction with the c-Kit and IL-7 signaling axis.

Guidance for RNA-seq co-expression network construction and analysis: safety in numbers

MOTIVATION

RNA-seq co-expression analysis is in its infancy and reasonable practices remain poorly defined. We assessed a variety of RNA-seq expression data to determine factors affecting functional connectivity and topology in co-expression networks.

RESULTS

We examine RNA-seq co-expression data generated from 1970 RNA-seq samples using a Guilt-By-Association framework, in which genes are assessed for the tendency of co-expression to reflect shared function. Minimal experimental criteria to obtain performance on par with microarrays were >20 samples with read depth >10 M per sample. While the aggregate network constructed shows good performance (area under the receiver operator characteristic curve ∼0.71), the dependency on number of experiments used is nearly identical to that present in microarrays, suggesting thousands of samples are required to obtain 'gold-standard' co-expression. We find a major topological difference between RNA-seq and microarray co-expression in the form of low overlaps between hub-like genes from each network due to changes in the correlation of expression noise within each technology.

CONTACT

jgillis@cshl.edu or sballouz@cshl.edu

SUPPLEMENTARY INFORMATION

Networks are available at: http://gillislab.labsites.cshl.edu/supplements/rna-seq-networks/ and supplementary data are available at Bioinformatics online.

Ribosomal and immune transcripts associate with relapse in acquired ADAMTS13-deficient thrombotic thrombocytopenic purpura

Approximately 40% of patients who survive acute episodes of thrombotic thrombocytopenic purpura (TTP) associated with severe acquired ADAMTS13 deficiency experience one or more relapses. Risk factors for relapse other than severe ADAMTS13 deficiency and ADAMTS13 autoantibodies are unknown. ADAMTS13 autoantibodies, TTP episodes following infection or type I interferon treatment and reported ensuing systemic lupus erythematosus in some patients suggest immune dysregulation. This cross-sectional study asked whether autoantibodies against RNA-binding proteins or peripheral blood gene expression profiles measured during remission are associated with history of prior relapse in acquired ADAMTS13-deficient TTP. Peripheral blood from 38 well-characterized patients with autoimmune ADAMTS13-deficient TTP in remission was examined for autoantibodies and global gene expression. A subset of TTP patients (9 patients, 24%) exhibited a peripheral blood gene signature composed of elevated ribosomal transcripts that associated with prior relapse. A non-overlapping subset of TTP patients (9 patients, 24%) displayed a peripheral blood type I interferon gene signature that associated with autoantibodies to RNA-binding proteins but not with history of relapse. Patients who had relapsed bimodally expressed higher HLA transcript levels independently of ribosomal transcripts. Presence of any one potential risk factor (ribosomal gene signature, elevated HLA-DRB1, elevated HLA-DRB5) associated with relapse (OR = 38.4; p = 0.0002) more closely than any factor alone or all factors together. Levels of immune transcripts typical of natural killer (NK) and T lymphocytes positively correlated with ribosomal gene expression and number of prior episodes but not with time since the most recent episode. Flow cytometry confirmed elevated expression of cell surface markers encoded by these transcripts on T and/or NK cell subsets of patients who had relapsed. These data associate elevated ribosomal and immune transcripts with relapse history in acquired, ADAMTS13-deficient TTP.

Epigenome profiling reveals significant DNA demethylation of interferon signature genes in lupus neutrophils

Recent evidence suggests that neutrophils play an important role in the pathogenesis of lupus. The goal of this study was to characterize the epigenetic architecture, by studying the DNA methylome, of neutrophils and low density granulocytes (LDGs) in lupus patients. We studied 15 lupus patients and 15 healthy age, sex, and ethnicity matched controls. Genome-wide DNA methylation was assessed using the Illumina HumanMethylation 450 BeadChip array, which includes over 485,000 methylation sites across the entire genome. Bisulfite DNA sequencing was used to validate the array results. Statistical and bioinformatic analysis was performed to identify and characterize differentially methylated loci and genes. We identified 293 differentially methylated CG sites in neutrophils between lupus patients and controls. The majority (68%) of differentially methylated CG sites were hypomethylated in lupus neutrophils compared to controls, suggesting overall hypomethylation. We found a robust and consistent demethylation of interferon signature genes in lupus neutrophils, and similar demethylation in the same genes in autologous LDGs. Indeed, the DNA methylome in lupus neutrophils and LDGs was almost identical, suggesting similar chromatin architecture in the two granulocyte subsets. A notable exception was the hypomethylation of a CG site in the promoter region of the cytoskeleton-regulating gene RAC1 in LDGs. Our findings demonstrate a pattern of robust demethylation of interferon signature genes in lupus patients supporting a pathogenic role for neutrophils in lupus. We suggest a model whereby DNA from lupus neutrophils and LDGs externalized by NETosis enhance type-I IFN production via TLR-9 stimulation by hypomethylated DNA.

GEM2Net: from gene expression modeling to -omics networks, a new CATdb module to investigate Arabidopsis thaliana genes involved in stress response

CATdb (http://urgv.evry.inra.fr/CATdb) is a database providing a public access to a large collection of transcriptomic data, mainly for Arabidopsis but also for other plants. This resource has the rare advantage to contain several thousands of microarray experiments obtained with the same technical protocol and analyzed by the same statistical pipelines. In this paper, we present GEM2Net, a new module of CATdb that takes advantage of this homogeneous dataset to mine co-expression units and decipher Arabidopsis gene functions. GEM2Net explores 387 stress conditions organized into 18 biotic and abiotic stress categories. For each one, a model-based clustering is applied on expression differences to identify clusters of co-expressed genes. To characterize functions associated with these clusters, various resources are analyzed and integrated: Gene Ontology, subcellular localization of proteins, Hormone Families, Transcription Factor Families and a refined stress-related gene list associated to publications. Exploiting protein–protein interactions and transcription factors-targets interactions enables to display gene networks. GEM2Net presents the analysis of the 18 stress categories, in which 17 264 genes are involved and organized within 681 co-expression clusters. The meta-data analyses were stored and organized to compose a dynamic Web resource.

Microarray meta-analysis of RNA-binding protein functions in alternative polyadenylation

Alternative polyadenylation (APA) is a post-transcriptional mechanism to generate diverse mRNA transcripts with different 3′UTRs from the same gene. In this study, we systematically searched for the APA events with differential expression in public mouse microarray data. Hundreds of genes with over-represented differential APA events and the corresponding experiments were identified. We further revealed that global APA differential expression occurred prevalently in tissues such as brain comparing to peripheral tissues, and biological processes such as development, differentiation and immune responses. Interestingly, we also observed widespread differential APA events in RNA-binding protein (RBP) genes such as Rbm3, Eif4e2 and Elavl1. Given the fact that RBPs are considered as the main regulators of differential APA expression, we constructed a co-expression network between APAs and RBPs using the microarray data. Further incorporation of CLIP-seq data of selected RBPs showed that Nova2 represses and Mbnl1 promotes the polyadenylation of closest poly(A) sites respectively. Altogether, our study is the first microarray meta-analysis in a mammal on the regulation of APA by RBPs that integrated massive mRNA expression data under a wide-range of biological conditions. Finally, we present our results as a comprehensive resource in an online website for the research community.

Bias tradeoffs in the creation and analysis of protein-protein interaction networks

Networks constructed from aggregated protein-protein interaction data are commonplace in biology. But the studies these data are derived from were conducted with their own hypotheses and foci. Focusing on data from budding yeast present in BioGRID, we determine that many of the downstream signals present in network data are significantly impacted by biases in the original data. We determine the degree to which selection bias in favor of biologically interesting bait proteins goes down with study size, while we also find that promiscuity in prey contributes more substantially in larger studies. We analyze interaction studies over time with respect to data in the Gene Ontology and find that reproducibly observed interactions are less likely to favor multifunctional proteins. We find that strong alignment between co-expression and protein-protein interaction data occurs only for extreme co-expression values, and use this data to suggest candidates for targets likely to reveal novel biology in follow-up studies.

BIOLOGICAL SIGNIFICANCE

Protein-protein interaction data finds particularly heavy use in the interpretation of disease-causal variants. In principle, network data allows researchers to find novel commonalities among candidate genes. In this study, we detail several of the most salient biases contributing to aggregated protein-protein interaction databases. We find strong evidence for the role of selection and laboratory biases. Many of these effects contribute to the commonalities researchers find for disease genes. In order for characterization of disease genes and their interactions to not simply be an artifact of researcher preference, it is imperative to identify data biases explicitly. Based on this, we also suggest ways to move forward in producing candidates less influenced by prior knowledge. This article is part of a Special Issue entitled: Can Proteomics Fill the Gap Between Genomics and Phenotypes?

Experimental validation of 5 in-silico predicted glioma biomarkers

BACKGROUND

Glioblastoma multiforme (GBM) is a high-grade glioma with poor prognosis. Identification of new biomarkers specific to GBM could help in disease diagnosis. We have developed and validated a bioinformatics method to predict proteins likely to be suitable as glioma biomarkers via a global microarray meta-analysis to identify uncharacterized genes consistently coexpressed with known glioma-associated genes.

METHODS

A novel bioinformatics method was implemented called global microarray meta-analysis, using approximately 16,000 microarray experiments to identify uncharacterized genes consistently coexpressed with known glioma-associated genes. These novel biomarkers were validated as proteins highly expressed in human gliomas varying in tumor grades using immunohistochemistry. Glioma gene databases were used to assess delineation of expression of these markers in varying glioma grades and subtypes of GBM.

RESULTS

We have identified 5 potential biomarkers-spondin1, Plexin-B2, SLIT3, fibulin-1, and LINGO1-that were validated as proteins highly expressed on the surface of human gliomas using immunohistochemistry. Expression of spondin1, Plexin-B2, and SLIT3 was significantly higher (P < .01) in high-grade gliomas than in low-grade gliomas. These biomarkers were significant discriminators in grade IV gliomas compared with either grade III or II tumors and also distinguished between GBM subclasses.

CONCLUSIONS

This study strongly suggests that this type of bioinformatics approach has high translational potential to rapidly discern which poorly characterized proteins may be of clinical relevance.

ELTD1, a potential new biomarker for gliomas

BACKGROUND

Glioblastoma multiforme (GBM), a high-grade glioma, is characterized by being diffuse, invasive, and highly angiogenic and has a very poor prognosis. Identification of new biomarkers could help in the further diagnosis of GBM.

OBJECTIVE

To identify ELTD1 (epidermal growth factor, latrophilin, and 7 transmembrane domain-containing protein 1 on chromosome 1) as a putative glioma-associated marker via a bioinformatic method.

METHODS

We used advanced data mining and a novel bioinformatics method to predict ELTD1 as a potential novel biomarker that is associated with gliomas. Validation was done with immunohistochemistry, which was used to detect levels of ELTD1 in human high-grade gliomas and rat F98 glioma tumors. In vivo levels of ELTD1 in rat F98 gliomas were assessed using molecular magnetic resonance imaging.

RESULTS

ELTD1 was found to be significantly higher (P = .03) in high-grade gliomas (50 patients) compared with low-grade gliomas (21 patients) and compared well with traditional immunohistochemistry markers including vascular endothelial growth factor, glucose transporter 1, carbonic anhydrase IX, and hypoxia-inducible factor 1α. ELTD1 gene expression indicates an association with grade, survival across grade, and an increase in the mesenchymal subtype. Significantly high (P < .001) in vivo levels of ELTD1 were additionally found in F98 tumors compared with normal brain tissue.

CONCLUSION

Results of this study strongly suggests that associative analysis was able to accurately identify ELTD1 as a putative glioma-associated biomarker. The detection of ELTD1 was also validated in both rodent and human gliomas and may serve as an additional biomarker for gliomas in preclinical and clinical diagnosis of gliomas.

Differential expression analysis for pathways

Life science technologies generate a deluge of data that hold the keys to unlocking the secrets of important biological functions and disease mechanisms. We present DEAP, Differential Expression Analysis for Pathways, which capitalizes on information about biological pathways to identify important regulatory patterns from differential expression data. DEAP makes significant improvements over existing approaches by including information about pathway structure and discovering the most differentially expressed portion of the pathway. On simulated data, DEAP significantly outperformed traditional methods: with high differential expression, DEAP increased power by two orders of magnitude; with very low differential expression, DEAP doubled the power. DEAP performance was illustrated on two different gene and protein expression studies. DEAP discovered fourteen important pathways related to chronic obstructive pulmonary disease and interferon treatment that existing approaches omitted. On the interferon study, DEAP guided focus towards a four protein path within the 26 protein Notch signalling pathway.

The data deluge represents a growing challenge for life sciences. Within this sea of data surely lie many secrets to understanding important biological and medical systems. To quantify important patterns in this data, we present DEAP (Differential Expression Analysis for Pathways). DEAP amalgamates information about biological pathway structure and differential expression to identify important patterns of regulation. On both simulated and biological data, we show that DEAP is able to identify key mechanisms while making significant improvements over existing methodologies. For example, on the interferon study, DEAP uniquely identified both the interferon gamma signalling pathway and the JAK STAT signalling pathway.